黑盒测试主要是根据产品的外部功能来规划测试,检查程序各个功能是否实现,主要的质量属性是否达到要求,其中有无错误。
暴力搜索优化的一个替代方案是黑盒(Black-Box)非凸优化技术。黑盒非凸优化算法可根据某些预定义的度量找到足够最佳的局部最小值(或最大值)的次优解。 Python具有许多这样的工具。
Procedural Noise Adversarial Examples for Black-Box Attacks on Deep Neural Networks论文笔记 0....Goodfellow, “Transferability in Machine Learning: From Phenomena to Black-box Attacks using Adversarial...Swami, “Practical Black-box Attacks Against Machine Learning,” in Asia Conference on Computer and Communications...Bethge, “Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models...Hsieh, “Zoo: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training
该文章提出一种利用程序化噪声来生成对抗样本的方法, 所提出的方法和那些通过梯度不断修改以至于到达分类器的边界的方法不一样, 上述方法需要对目标的模型有一定的了解...
本文介绍NeurIPS 2020接收论文《Learning Black-Box Attackers with Transferable Priors and Query Feedback》。 ?...基于这一观察,我们提出了可学习的黑盒攻击方法(Learnable Black-Box Attack,LeBA)。目前,鲜有使用查询反馈来更新替代模型的研究。...Improving black-box adversarial attacks with a transfer-based prior....Simple Black-box Adversarial Attacks. ICML 2019. [4] Guo Y, Yan Z, Zhang C....Square attack: a query-efficient black-box adversarial attack via random search.
We apply AdvGAN in both semi-whitebox and black-box attack settings....In black-box attacks, we dynamically train a distilled model for the black-box model and optimize the...Our attack has placed the first with 92.76% accuracy on a public MNIST black-box attack challenge. ?
Segmentation Poster 2274 Adversarial Ranking Attack and Defense Poster 2336 Boosting Decision-based Black-box...Adversarial Perturbations Using Context Inconsistency Poster 4362 Square Attack: a query-efficient black-box...Bi-directional Likelihood Regularization for Visual Classification Poster 4889 Improving Query Efficiency of Black-box
原文标题:Some Black-box Reductions for Objective-robust Discrete Optimization Problems Based on their LP-Relaxations
Addressing Unmeasured Confounder for Recommendation with Sensitivity Analysis [6] Knowledge-enhanced Black-box...to the security and privacy concerns, it is more practical to perform adversarial attacks under the black-box...However, generating high-quality fake user profiles under black-box setting is rather challenging with...More specifically, we propose a knowledge graph-enhanced black-box attacking framework (KGAttack) to...integrated into hierarchical policy networks to generate fake user profiles for performing adversarial black-box
a Service-Oriented Architecture. 4 基于监控指标的异常检测 4.1 异常检测 4.1.1 无监督检测 Detecting Anomalous Behavior of Black-Box...FChain: Toward Black-Box Online Fault Localization for Cloud Systems. 4.2.2 基于拓扑图的分析 Graph-based root
Zoo: Zeroth order optimization based black-box attacks to deep neural networks without training substitute...Decision-based adversarial attacks: Reliable attacks against black-box machine learning models[J]. arXiv
文题中的“黑”与“白”指的是神经网络的black-box的知识表达和我希望的white-box的可解释性模型。...作为博后研究员,我在UCLA带领几个学生探索一种语义层面可解释的神经网络特征表达,或者把black-box网络特征表达转化成white-box图模型。...▌Disentangling CNN representations into graphical models ---- 其实在black-box表达中,之前提到的representation bias...但是让人尴尬的是,恰恰是神经网络black-box表达保证了特征提取的flexibility和信息表达的效率(见information bottleneck理论)。
相关论文已经被KDD 2021录用,"OpenBox: A Generalized Black-box Optimization Service"。..."OpenBox: A Generalized Black-box Optimization Service."..."Google vizier: A service for black-box optimization."
2209.07923 代码/Code: https://github.com/BGU-CS-VIL/DeepMCBM A Large-scale Multiple-objective Method for Black-box
Functionaltesting(功能测试),也称为behavioral testing(行为测试)、黑盒测试或数据驱动测试 黑盒测试(Black-box Testing,又称为功能测试或数据驱动测试
2204.00367 代码/Code: https://github.com/lingyanruan/DRBNet Investigating Top-k White-Box and Transferable Black-box
TD3 Applications 9 Model-Based RL 10 Imitation Learning with the DAgger Algorithm 11 Understanding Black-Box
On one hand, black-box models need to be made white; for this there are many efforts, e.g., trying to...improve the comprehensibility of black-box models [8], or even produce accurate and comprehensible models
最佳论文奖由斯坦福大学Pang Wei Koh和Percy Liang的论文Understanding Black-box Predictions via Influence Functions获得。...量子位整理传送如下: 最佳论文奖 论文:Understanding Black-box Predictions via Influence Functions(通过影响函数理解黑箱预测问题) 作者:Pang
Representations 论文/Paper: http://arxiv.org/pdf/2205.04175 代码/Code: None Bandits for Structure Perturbation-based Black-box
领取专属 10元无门槛券
手把手带您无忧上云